Adaptive multivariate fault detection

Abstract

A method and apparatus for detecting faults. A set of data samples is received, the set of data samples including multiple process variables. One or more multivariate statistical models are adapted, wherein adapting includes applying a change to at least one univariate statistic of the one or more multivariate statistical models if the change is greater than a threshold value. The one or more multivariate statistical models are used to analyze subsequent process data to detect faults.

Description

[0001] related application [0002] This application claims priority to Provisional Application No. 60 / 746649, filed May 7, 2006, and Provisional Application No. 60 / 746647, filed May 7, 2006. technical field [0003] Particular embodiments of the present invention relate to error detection, and more particularly to error detection using multiple error signatures. Background technique [0004] Many businesses employ sophisticated manufacturing equipment that includes multiple sensors and controls that are carefully monitored during processing to ensure product quality. One method of monitoring these multiple sensors and controllers is statistical process monitoring (a means of performing statistical analysis on sensor measurements and process control values ​​(process variables)), which enables automatic detection and / or error detection . A "fault" can be a malfunction or misalignment of manufacturing equipment (such as a deviation of an operating parameter of a machine fro...

AI Technical Summary

Problems solved by technology

However, by adjusting only univariate means and scaling coefficients, this method cannot provide adjustment of these covariances among variables within a model

[0011] Each of the common adaptation methods described above is susceptible to cumulative computational rounding errors caused by the periodic adaptation

This then results in the model having imprecise statistics which can cause both false alarms and failures to detect errors

Images